Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

One thing that catches my eye here is the use of a smart plug on the refrigerator for current monitoring. I've tried a ton of different ones including the "good" ones like Shelly, and they all seem to use shunt resistors to gauge power draw. It would make me really nervous to use something like that to measure power draw on a big inductive load like a fridge. It's a shame, I've never seen a current clamp in plug form with no on/off switch, so you've basically gotta do some fab work, but that's basically really the only safe way to collect current data for anything that pulls a non-trivial amount of power.


Can you elaborate a bit further on what you’d need to do and why? It’s been a while since my electrical courses.

I’ve been trying to measure home power consumption with these plugs (and the ones from IKEA) but I’ve been getting suspicious readings for inductive loads.


I wouldn't bother using a smart plug. The other thing that's real dangerous about them that I didn't really elaborate on is the relay. There's a reason on something like a pump, fridge, AC unit, etc. you'd see a real contactor instead of a relay, and it's because relays are inappropriate and have dangerous failure modes for big loads like that and are typically way too small.

In most real non-resi situations, you'd probably isolate the hot leg and put a good CT (current clamp) on it and read that. The great thing about that is you haven't added anything in the power path for the device, like a shunt which is what most smart plugs use. Current clamps are good for a lot more current (though I guess a proper shunt could do it too). The easiest way to do this in your setting is to find a good UL-listed electrical box with cable glands, a short piece of DIN rail, a male and female plug pigtail, some proper THHN wire and wirenets and a Shelly 50A EM Pro, and just graft the EM Pro into the box and wire it up with it's CT. You've now got something signficantly more durable and probably safer (and correctly specced for the load). I've done other things like using an HV Labjack and some good CTs or other one or few off designs. There's lots of stuff in the commercial/industrial space that does this well but it tends to be $$$. Again, for the sake of my own family, I wouldn't use non-UL stuff (most plugs and things that go in gangboxes that are "smart" aren't UL listed, and MAYBE are ETL) as you who knows how much or how well it's tested.


The IKEA plugs have a tidbit in the manual that says a "motor load" is limited to 300W while a resistive load can reach the full 3680W (at 230V, probably less if you're in a lower voltage country).

Should be fine for modern fridges but older fridges may overwhelm the circuitry.


Isn't this just a question of correctly sizing the shunt?

Hall effect would also be possible, but more expensive.


Yeah, but you don't get do the sizing, it's done by whatever the engineer that designed the plug wanted. I don't know why you'd do hall effect for an AC circuit, a good regular old current clamp should get you as much accuracy as you'd ever want.


Just to be pedantic, you could have a current meter where the clamp/hall sensor is internal and not clamped on the power cord.


That's what I'm say, but I've never ever seen one. I would love to buy one with a current clamp integrated, as it would save me from needing to fab a box and build something, but I don't know of any that exist today.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: